What is a primary downside of using Spot Instances for workloads?

What is a primary downside of using Spot Instances for workloads?

• A. They require long-term commitment
• B. They have constant availability
• C. There is no guarantee of instance availability
• D. They cost more than On-Demand

Answer: (C)

Spot Instances are cheap spare capacity, but they can be reclaimed by the cloud provider whenever capacity is needed. That means there is no guarantee that the same instance will be available for your workload for any given period. This interruption risk makes them unsuitable for critical, long-running, or stateful tasks unless you architect for interruptions (e.g., using fault tolerance, auto-scaling with mixes of instance types, and quick recovery). Why the other statements don’t fit: Spot Instances are not about long-term commitment; they’re the opposite—short-lived by design. They don’t provide constant availability, since they can be interrupted, which is the very reason for the lack of guarantee. They typically cost less than On-Demand, not more, so pricing isn’t the downside.

Spot Instances are cheap spare capacity, but they can be reclaimed by the cloud provider whenever capacity is needed. That means there is no guarantee that the same instance will be available for your workload for any given period. This interruption risk makes them unsuitable for critical, long-running, or stateful tasks unless you architect for interruptions (e.g., using fault tolerance, auto-scaling with mixes of instance types, and quick recovery).

Why the other statements don’t fit: Spot Instances are not about long-term commitment; they’re the opposite—short-lived by design. They don’t provide constant availability, since they can be interrupted, which is the very reason for the lack of guarantee. They typically cost less than On-Demand, not more, so pricing isn’t the downside.